The American View: When the Nest Empties—but the Cables Stay

Consider that the academic models for how to run a successful business assume that most companies’ IT infrastructure and equipment are cost centres, not profit centres. That is, they’re necessary costs of doing business but don’t generate products or profits on their own. Therefore, a “smart” company should only purchase what the workers need to perform their core tasks and nothing more! Don’t spring for £100 ergonomic keyboards if disposable plastic £10 models will do. Don’t buy large monitors to lessen eye strain if small, fuzzy monitors will do. The network servers only have to work … not to work well, etc.

Additionally, it’s considered a best practice to keep expenses low by not stocking unused hardware or spare parts. Live that just-in-time logistics life instead! Why have the tools or parts you might need ready on a shelf when you can buy only what you need right after you need it? It’s not like you’ll ever find yourself in a technology-based emergency that can’t be put off for a few days …

Of course, as a former IT support person I can attest that this expense-minimization-focused academic model is suicidal nonsense. Sure, it keeps a “cost centre’s” budget low, but it also exposes the organisation to unnecessary operational risk. Minimal IT investment is only a competitive advantage so long as nothing ever goes wrong. Ask your help desk staff how many days they’ve gone without experiencing a drop-dead emergency in the last year. Go on … I’ll wait. [2]

This, I argue, is why us old school IT support types are famous for maintaining unofficial (and often hidden) stashes of operational spares, repair parts, cables, and adapters. When a busy executive’s laptop shorts out, they want a working replacement right now, not in 2-3 business days (weather permitting). One of the techniques we used in 2000s to address this risk was to keep a few laptop hard drives running out gold master image at the help desk walk-up service window. Whenever a user’s laptop went tango uniform, we’d simply remove the bad drive, pop in a new one, connect the “new” PC to the domain and send the user back to work in less than five minutes. Since all their apps were preloaded on the standard image and all the users’ documents were safely stored on the network, no work was lost. It was a small expense for parts and small investment of time to prepare that prevented tons of heartburn. 

Additionally, we deliberately bought gear that allowed for growth and flexibility. We made sure that all our new PCs could be upgraded with more RAM, larger hard drives, popular peripherals, and the like. Once a user needed “more computer,” we already have the parts on hand to “improve” their PC on request. It was much cheaper than buying a whole new computer, but – more importantly – much faster than waiting on procurement. Of course we kept the old parts for spares; some RAM sticks might serve three or four times in different cases before they were retired.